End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement Learning (2026.acl-long)
Copied to clipboard
| Challenge: | Existing multi-agent reinforcement learning methods depend on large critic networks to evaluate joint actions, leading to instability and high memory costs. |
| Approach: | They propose a method to optimize large language models for agent-specific roles . they propose combining agent-based frameworks with retrieval-augmented generation . |
| Outcome: | Experiments show that multi-agent group policy optimization outperforms baselines in task performance and computational efficiency. |
Similar Papers
Empowering Multi-Turn Tool-Integrated Agentic Reasoning with Group Turn Policy Optimization (2026.acl-long)
Copied to clipboard
Yifeng Ding, Hung Le, Songyang Han, Kangrui Ruan, Zhenghui Jin, Varun Kumar, Zijian Wang, Anoop Deoras
| Challenge: | Current reinforcement learning methods suffer from coarse-grained, trajectory-level rewards that provide insufficient learning signals for complex multi-turn interactions, leading to training stagnation. |
| Approach: | They propose a novel RL algorithm for training large language models for multi-turn tool-integrated reasoning (TIR) that incorporates three innovations: turn-level reward assignment that provides fine-grained feedback for individual turns, return-based advantage estimation where normalized discounted returns are calculated as advantages, and self-supervised reward shaping that exploits self-supervision signals from generated code to densify sparse binary outcome-based rewards. |
| Outcome: | The proposed algorithm outperforms GRPO by 3.0% across diverse math reasoning benchmarks and improves grepo by 3.9% on commonsense reasoning and program synthesis tasks. |
DPEPO: Diverse Parallel Exploration Policy Optimization for LLM-based Agents (2026.acl-long)
Copied to clipboard
JunShuo Zhang, Chengrui Huang, Feng Guo, Zihan Li, Ke Shi, Menghua Jiang, Jiguo Yu, Shuo Shang, Shen Gao
| Challenge: | Existing approaches to large language model (LLM) agents that follow the sequential "reason-then-act" paradigm suffer from limited exploration and incomplete environmental understanding as they interact with only a single environment per step. |
| Approach: | They propose a paradigm that enables an agent to interact with multiple environments simultaneously and share cross-trajectory experiences. |
| Outcome: | The proposed paradigm achieves state-of-the-art (SOTA) success rates while maintaining comparable efficiency to strong sequential baselines. |
Towards Effective and Efficient Multi-Agent Language Model Systems: Foundations, Prospects, and Applications (2026.acl-tutorials)
Copied to clipboard
| Challenge: | Multi-agent systems powered by large language models still face challenges . tutorial focuses on three core components to build effective and efficient systems . |
| Approach: | This tutorial introduces recent advances in building effective and efficient multi-agent LLM systems . it focuses on three core components: model distillation, dynamic routing, memory- and compute efficient serving . |
| Outcome: | This tutorial introduces state-of-the-art techniques for building efficient and efficient multi-agent LLM systems . it covers coordination and communication among agents, crucial for collective performance . |
Towards Efficient LLM Grounding for Embodied Multi-Agent Collaboration (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for grounding large language models suffer from inefficient querying . Existing approaches that rely on physical verification or self-reflection suffer from excessive querying. |
| Approach: | They propose a framework that introduces Reinforced Advantage feedback for efficient self-refinement of plans. |
| Outcome: | The proposed framework surpasses baselines in success rate and significantly decreases interaction steps of agents and query rounds of LLMs. |
LEPO: Latent Reasoning Policy Optimization for Large Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing latent reasoning methods that use chain of thought (CoT) are limited to selecting one discrete token at each reasoning step, which potentially induces information loss. |
| Approach: | They propose a framework that injects controllable stochasticity into latent reasoning via Gumbel-Softmax, restoring LLMs' exploratory capacity and enhancing their compatibility with Reinforcement Learning (RL). |
| Outcome: | The proposed framework preserves richer information for more comprehensive reasoning and is compatible with Reinforcement Learning (RL). |
Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration (2026.acl-long)
Copied to clipboard
| Challenge: | Existing context window extension methods obstruct scaling external knowledge input. |
| Approach: | They develop a multi-agent framework to overcome two core bottlenecks in existing agent orchestration designs. |
| Outcome: | The proposed framework overcomes two core bottlenecks and improves inference-time knowledge integration without longer-context training. |
RLAE: Reinforcement Learning-Assisted Ensemble for LLMs (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing ensemble methods for ensembling large language models rely on fixed weighting strategies that fail to adapt to dynamic, context-dependent characteristics of LLMs. |
| Approach: | They propose a framework that reformulates LLM ensemble through a Markov Decision Process. |
| Outcome: | The proposed framework outperforms existing methods by 3.3% on a diverse set of tasks while achieving lower time latency. |
AgentGym: Evaluating and Training Large Language Model-based Agents across Diverse Environments (2025.acl-long)
Copied to clipboard
Zhiheng Xi, Yiwen Ding, Wenxiang Chen, Boyang Hong, Honglin Guo, Junzhe Wang, Xin Guo, Dingwen Yang, Chenyang Liao, Wei He, Songyang Gao, Lu Chen, Rui Zheng, Yicheng Zou, Tao Gui, Qi Zhang, Xipeng Qiu, Xuanjing Huang, Zuxuan Wu, Yu-Gang Jiang
| Challenge: | Large language models (LLMs) are promising foundations to build generally-capable agents . however, the community lacks a unified interactive framework that covers diverse environments for comprehensive evaluation of agents. |
| Approach: | They propose a framework that features 7 real-world scenarios, 14 environments, and 89 tasks for unified, real-time, and concurrent agent interaction. |
| Outcome: | The proposed framework features 7 real-world scenarios, 14 environments, and 89 tasks for unified, real-time, and concurrent agent interaction. |
Enhancing Efficiency and Exploration in Reinforcement Learning for LLMs (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches allocate an equal number of rollouts to all questions during the RL process, which is inefficient. |
| Approach: | They propose a mechanism for dynamically allocating rollout budgets based on the difficulty of the problems, enabling more efficient RL training. |
| Outcome: | The proposed model improves response precision while preserving exploratory ability to uncover potential correct pathways. |
Group-Aware Reinforcement Learning for Output Diversity in Large Language Models (2025.emnlp-main)
Copied to clipboard
Oron Anschel, Alon Shoshan, Adam Botach, Shunit Haviv Hakimi, Asaf Gendler, Emanuel Ben Baruch, Nadav Bhonker, Igor Kviatkovsky, Manoj Aggarwal, Gerard Medioni
| Challenge: | Large Language Models suffer from mode collapse, repeatedly generating the same few completions even when many valid answers exist. |
| Approach: | They propose a group-aware policy optimization extension of GRPO that computes rewards over the group as a whole. |
| Outcome: | The proposed model improves on standard LLM benchmarks without compromising accuracy. |